There are many factors that may make comparison among microphones difficult. In addition to human factors, there are just so many variables in the equipment configuration. I have used numerous microphones with a switchable microphone mixer having a line output and calibrated input levels. I believe this system is optimized to show the difference, if there is any, among microphones. Details of a microphone mixer-based multi-microphone system are described here:
http://www.voicerecognition.com/board/inde...=18463&hl=zmgaoThe truth is, at a certain level the microphone system starts to demonstrate maximized accuracy where all microphones above that level are just equally accurate. In addition to actual voicerecognition experience, I also made voice recordings and actually listened to them using a high quality headphone. I am convinced that most professional microphones are equally good because all recordings sound so clean and clear to my ears that it is hard to imagine different microphones would make any difference with a rather primitive speech engine we have now.
Note that I used the term "at a certain level", which in my experience means a quality professional microphone. I am by no means arguing that all microphones are made equal for voicerecognition. But on the other hand, I do not believe the myth that among the high-quality microphones, there are only just a special few that are particularly good for voicerecognition.
The bottom line: in terms of microphone quality, I have total peace of mind now, knowing that I have the best already. There is absolutely no way I could improve my accuracy by buying another microphone. The problem, if there is any, remains with the software.
By the way, I have long believed that the large variance of microphone performance reported has a lot to do with variations in the connectivity and connection conditions. One may get a top-quality microphone but use it in a poor configuration and claim that the microphone is not good. The fact that many people report excellent performance of some USB microphones and Bluetooth microphones is evidence to support my point. Compared to professional microphones, these have a relatively poor microphone element and a very low-grade analog to digital converter, so it makes you wonder about the superior performance experienced by people. It's not that people are making false reports. It is just that with USB and Bluetooth there is very little to go wrong with the connections and configurations and very little to deviate from an optimized voice level, and as a result statistically more people tend to see better results (or more precisely, fewer people experience unknown problems).
If you have the knowledge to set up a high-quality system, you may have every reason to be confident that Sennheiser ME-3 is a better microphone than all other consumer grade microphones, including the Plantronics CS-55. Because there are so many variables involved, following other people's opinions in choosing a microphone is difficult. The best to do of course is to test it by yourself. But if that is not possible, just go after the microphone that is recognized to have the best microphone element. Given proper connection and configuration, it is the quality of the microphone element that matters. In the example of Sennheiser ME-3 and Plantronics CS-55, the former is a $150 microphone that has at least $100 value going into the microphone element, while the latter is a $250 microphone that has less than $50 in the microphone element and the rest in the other technology, which I have no doubt is a great value to many people but simply could not contribute to accuracy.